A description of a pattern across six tagged crashes — not a diagnosis, not a predictor, and not a rule for what happens next. The rest of this report walks through exactly how we got here, signal by signal — what's helped other people is in your full profile.
Where your day tends to go (from your wearable, median minutes):
About 38.1% of your sedentary time reads as a 'stress' state by the tracker (computed from your daily stress-share, not the ratio of the three minute-totals above — the two won't match exactly). These are medians per category, not a full split of 24 hours: some time each day isn't classified into any category, and category minutes can vary independently.
The makeup of your typical night (median minutes):
Your night is about 6.3 h of sleep (plus about 23 min awake) — this total is its own nightly median, not a sum of the stage minutes above, so it won't match them exactly. Sleep numbers don't always capture how rested you feel — that's okay.
Resting heart rate around 64 bpm. Your HRV is shown only as your own trend, not compared to others.
These are not 'steps' — they're the deeper physiological signals and next-day patterns we researched specifically for conditions like yours (a mix of what your body measured and what you reported). For each: what it is, what we know about it, and how it behaved for you over time.
How to read each chart: the bold line is that measure (for example your cardiac cost or recovery completeness), smoothed over 7 days; the grey dashed line is your usual level (a rolling baseline); the blue band is your usual range; red marks are your crashes. When the bold line moves out of your usual band in the days just before a red mark, that measure may be an early sign. Headline numbers like "~20 bpm" come from your raw daily values around each crash, not the smoothed line — smoothing is only there to make the shape easier to see.
The verdict word on each signal (read this first): leans — moved in a concerning direction before your crashes; maybe — a weak lean that doesn't clear the bar; quiet — no steady link visible yet (that's an answer too); unclear — the statistical and the raw signals disagree, so we don't call it. All descriptive, over 6 episodes — leads to watch, not proven predictors.
Here are two of the six signals we track, starting with your clearest.
All episodes overlaid and averaged. On each strip: the grey band is your usual range (± typical spread), the orange dot is the days just before a crash, the red diamond is the crash day.
All episodes overlaid and averaged. Cardiac cost rises above its usual through the run-up to a dip, while your steps stay near their usual — the path runs through your heart rate during activity.
On the days your cardiac cost stays in its usual range, a dip doesn't tend to follow. That's a pattern, not proof — staying in range might not be what prevents a dip, it could just as easily mark a calmer stretch overall. Worth watching, not a plan.
In your own notes, the days before a dip more often carry: over-exertion (the most frequent one), poor sleep, headache, fatigue, brain fog. Your heart-rate data doesn't contradict that — it's not that your steps and overall load changed, it's that the same effort cost more. These may be the same thing, seen two ways. Your own observations on small numbers — leads to watch, not verdicts.
Your day-by-day load (steps and MET) in the days around a crash — read it alongside your headline signal above. Load itself often drops on the crash day, once the dip has already started.
A bit of context from science and our knowledge base, applied to your picture:
There's more in your full profile. Community members get the complete picture: all six deep signals checked one by one — not just the two featured here — the rhythm between your crashes and which days tend to carry more sedentary stress, specific things that have helped other people with a similar pattern, and how your own "type" compares to others in the community working through a similar journey.
The full version is available to community members only.
Welltory Energy Lab · general information from community data, not a substitute for medical advice.
Provenance: labels episodes_v2_all_sources (Snapshot2); data lake day_info/hrv refreshed through 2026-08-02, minute_stream through 2026-07-13 — this specific analysis is fixed to your data through 2026-06-07 for reproducibility. Descriptive observations, not validated predictors (N=6 episodes). Educational layer — from KB + our science corpus (clinical items → Anna review). Group figures — k-anonymity, no cross-person HRV, no rankings.